Faculty Colloquium: Dr. Michael Melloch

Faculty Colloquium: Dr. Michael Melloch

October 10, 2019 0 By Stanley Isaacs


>>Good afternoon, everyone. My name is and for
those of you who don’t know me, I’m a professor and
head of the school of DCP. Now, it’s my pleasure to welcome you
to this celebration of faculty and careers colloquium. Typically the production of the speaker
is done by the associate Dean, or the Dean of the Dean’s office. But, occasionally the honor or the
privilege falls upon somebody like mean, and it’s my good fortune that I have
the pleasure of introducing Mike to you. Before I introduce Mike,
I’d like to introduce the program to you then finally we have 20, the celebration
of faculty careers comes out of our strategic plan initiatives and part of it
is something called the faculty of 2020. And then the other component of it which
is aligning of the promotion engineer policies with the goals of the college and
things like it, in similar terms here’s how I see it. Most of us, we know Mike as a colleague,
we see what he does. But occasionally it’s great for us to sit and
listen to his perspective of his career. So it’s a great way for
Us to learn something about Mike’s, that’s goal number one. Secondly, I think it’s good for
all of us, Senior Faculty especially, to occasionally sit back and take an expansive view of where the has
been, and find a for the future. So those are the two reasons
why I think we have these here. And again it’s my pleasure
to introduce Mike. Just in terms of writing off the project
itself, the celebration of faculty careers was piloted successfully in spring 2013
and its full implementation in 2013 14. [INAUDIBLE] Obviously it’s continuing. And now it’s my pleasure
to introduce Mike to you. Mike got his BSEE MSEE and
PHD all degrees.>From Purdue, so he got his bachelors and
masters in 1975 and 1976. Then he went away and
joined Intel Corporation, where he worked on the 8748,
the first single-chip microcomputer. And 8051, the second-generation
single-chip microcomputer. And then he came back and got his Ph.D. PHD in 1981 and then he went away again. So you see the theme,
you can see the connection here. He went away again and
then joined the Central Resource Labs at TI Texas Instruments as a member
of the technical staff. At TI his research interests centered
around gallium arsenate surface. Devices.
When August 1984 came back and Jonathan associate professor
associated for associate head of. He’s American Physics Society and Institute of Electrical and
Electronics Engineers, IEEE. The Optical Society of America,
the American Vacuum Society The same, but he’s also received the 2008,
2013 Motorola Excellence in Teaching Award in ACD,
but 2009 and 2014 Award for Teaching Excellence 2012
Outstanding Professor Award and the 2012 Murphy Outstanding
Undergraduate Teacher award, this is the highest teaching award given by
Purdue University for individual teachers. So he’s won that. And then college of
engineering of 2016 award for teaching and obviously you know
the best teacher at Purdue. It doesn’t make sense Win the Potter Award
and you won that in 2016. [LAUGH] To finish off he was also inducted
in 2012, in Purdue Teaching Academy. In 2013 he was inducted into
Purdue Book of Great Teachers. So I think of Mike’s career
as being made up of two acts. The first act was when he
was a stellar researcher. He’s pretty much a fellow of
every society you can think of. And then his second act is more recent,
I think, and that is he has focused very
much as a patron on teaching. And in fact, I think of him as
a conscience of teaching as the associate head, he gets to
interview every faculty [INAUDIBLE] in our teaching [INAUDIBLE] we may get
to a part where we [INAUDIBLE] or not. He is a great councilor for faculty members especially junior faculty
members about how they can be betters. In short he is the [INAUDIBLE] so
I’m very pleased and privileged to I served with him as
the head of associate head. So we’re gonna learn a lot more about
Mike, but please join me in welcoming him.>>[APPLAUSE]
>>Thank you, Ragu. Well, the hardest part about putting this
talk together was coming up with a title. And I had some that were
a lot worse than than.>>[LAUGH]
>>Okay, just to give you a brief background. I was born in Hammond, Indiana. When I was about six we moved from
East Chicago to Highland, Indiana. Graduated from Highland High School,
my parents were young adults when the Great Depression hit,
so the Great Depression and World War II had a big influence on
them and hence a big influence on me. So, I went to Purdue because it was in state, and inexpensive, and
It was only 90 miles away. I’ve got my BS and MS and
s we do mention and I went and was a designer engineer
at Intel for a couple of years. Came back and got my PhD. Left and was at TI for three year and
then kept coming back. So, I came back for
I guess the third and final time. Okay so my father was in college
when the great depression hit so he had to drop out of school. He would have been the first college
graduate in either side of my family. But from my oldest recollection, my family stressed education. And I can still recall, and
I was less than six years old because this was at the kitchen
table in East Chicago, my grandmother talking to me about
education and going to college. So, I always had that in the. In the back of my mind. And my mom had two sisters and each one of the sisters
had a son who got a PhD. My two cousins who are about eight years
older than me, so they did it first. One got a PhD in Biochemistry
from Wisconsin and the other one a PhD in
Electrical Engineering from Cornell. [INAUDIBLE] So when World War II came, my dad went into the army,
and he served in Europe. He was a prisoner of war in Germany for
a while. And in fact, when he was free,
his weight had dropped under 100 pounds. So, he had health issues for
the rest of his life from that. In the army did rule Hamas
being disabled because of that. One nice thing about this program is that
I went and looked back at things and i discovered,
my dad didn’t talked about it much, but i discovered where he
was held as a POW So, it was always stressed about education,
going to college and one of my biggest worries was ever
being able to financially afford it. But my dad being a disabled veteran,
the state of Indiana had, well they still have a program for
children of disabled veterans. Where there’s a tuition and fee exemption. So when I came to Purdue,
the tuition was $350 a semester.>>[LAUGH]
>>But for me,
it was only $90 because of this program. So for 8 semesters,
$720 what it cost me to go to Purdue.>>[LAUGH] Now, when I originally applied,
I applied in the school of science and was gonna go into physics,
because I had never heard of engineering, I had no idea what engineering was about,
but after being accepted to Purdue,
I looked into it more, and I actually switched to engineering
before I actually showed up on campus. Which was third week in September,
cuz back then the semester started about the third week in September,
and it actually ended in January. You’d go home at Christmas winter or
break and then you’d come back for another week of classes and
then final exams. So it was also fortunate that
Purdue has everyone go in the first year engineering because I just
learned with engineering was about but I had no idea what discipline to go into,
so it was useful to be in a first year engineering program rather
than picking a discipline. And one other thing, back then I know
they still had these courses like we have a EE 190 but
it was completely different back then. That was a three credit hour course and
you really did things. You got into circuit design, power. So you really got a good flavor for
the discipline. So I got my Bachelors in 75 and
I didn’t think I could do anything. How could I go to a company and
really contribute? So I stayed for my Master’s. Also, back then, they didn’t have
these summer intern programs, during the summer, if you do something,
you worked at a restaurant or for a couple summers I worked
at the Parks Department. So I was cutting grass,
lining baseball fields. A couple summers I worked for
the Street Department and that also entailed being on the back
of a garbage truck sometimes. So if I had any doubts about
getting a college degree.>>[LAUGH]
>>So then with these internships, as I was working on my Masters, Dave Sam,
who’s one of our distinguished engineering alumni, was a fraternity brother of mine,
and he was working at Intel. So I called him up and said, is there
a chance I could come out there and work for the summer? And he said, well, let me talk to my boss. He called me back and
said, sure, come on out. So I drove from West Lafayette
to Santa Clara, California. And that’s when I discovered it was really
an interview It wasn’t an actual job, but->>[LAUGH]>>But think it was a formality. But so I talked to some engineers,
talked to Dave’s Boss, Hank Bloom. Talked to Hank Bloom’s boss,
Sean Claude, Cornet. And this is why I really know it’s
probably a formality, cuz Jean, even though I had a couple
years of French in high school. Jean Claude had this accent
that when he interviewed me. I could understand only one word and
he said it a lot though, microprocessor.>>[LAUGH]
>>So I just smiled and shook my head. So I worked during the summer
at Intel on the 8275 which uses CRT controller chip and Danny Joe
was the only engineer on that and so what I was doing for
him is the circuits were designed and I had to run simulations to size
all the transistors and I must have done a good job because when I left
that summer, I had a job offer in hand. So back then, two engineers would work designing a microprocessor or
microcomputer. The 8748 was the first microcomputer. And what that means is, it had program memory and
data memory on it. So it could standalone run a gas pump, washing machine, microwave oven. So here’s a [NOISE] plot of it. [NOISE] So
there’s an e prime across the top here. So here’s the e prom for your program
memory, here’s the data memory and all the logic circuit. When this came out,
it was the largest transistor of the time. The chip at the time,
it was 220 by 240 mills in diameter. Clear here, because it was a prom and
the way you erased it was with UV light. So the engineers who designed this
were Dave Stan and Dave Buddy. So I show up at Intel,
start being a permanent employee and Dave Stan had moved on to a peripheral
chip for the 8086 family. The 8089. Day Buddy was from Oregon and
he wanted to go back. And Intel had just opened
a place in Hillsboro and he said, you either send me there now or
I’m going somewhere else. So they were both gone, so here,
I said, half a year earlier I felt like I couldn’t do anything, I walk in
the door and about a week later I’m in charge of the 8748 and it still was
not a working functioning chip. So it had gone through what we refer to as
three stepping, so the A stepping B and C, so a stepping you process the chip,
it comes back, you test it. Find out as much as you can
about why it’s not working. Make the fixes, do another processing run. So we were down three runs like that,
and things still weren’t working. So my first job was
getting that chip working. And I’ll just tell you one story. I remember one thing we fixed. The previous summer,
when I started as my summer job, Bill Goodman started permanently. So he was fresh out of Kansas State. So by the time I went
back four months later, he was pretty seasoned
engineer back then Intel. He’d been there almost seven,
eight months.>>[LAUGH]
>>My boss, Gene Hill, was 30, and he was a real old timer. He was like Obi Won Kenobi.>>[LAUGH]
>>And that’s appropriate because Star Wars
came out while I was working at Intel. So anyway, Bill and I,
we were good friends. And so, one night we stayed up all night to try
to figure out something that was wrong. Because there were some instructions that, this chip was supposed to
run at like 4 megahertz and once you got it past like 1 megahertz,
things stop functioning. So we stayed up one night,
figured out which for the problem and
that product I was showing you what you We had to go through and by the end
I knew what everything on there was. I could identify every transistor and
why was there. But you have this little loops and
you would look at the layout and you could measure dimensions and
we found one transistor was drawn two microns,
the channel was two microns too long. So fixing that fixed the problem. I still remember it’s the ID
to port to lower signal.>>[LAUGH]
>>So after that happened,
then moved on to the 8051. And this might have been the last chip
at Intel where really two people did the whole thing. So I was codesigner with Steve Sample. That’s not the same Steve Sample
who was an EE professor here, and then was the president at
University of Southern California. This was a different Steve Sample. So each one of us had responsibility for
about half the chips. You would do the logic design,
then you would do the circuit design. Then, you would simulate the circuits and
size all the transistors. Then, you would oversee,
I forget what we call them. They weren’t called drafts people. But they were the people who would take
your circuit designs and dimensions, and draw all those lines and
things you saw in that plot I had earlier. And then, we had to oversee to make
sure everything was being drawn right. So in the middle of that process of,
everything was designed, except laying out the mask I
decided to come back and get a PhD. So I came back and when I was doing my
masters I had TA’d this course 190 so I TA’d it again, and
what I’ve done at Intel was a kinda mixture of solid state and
computer engineering. So I really didn’t know which
direction I wanted to go. But I took a course the first semester
called Surface Acoustic Wave Devices that Bob Gunshor was teaching. And after the first exam,
he had written, see me, on the exam. So I thought I was in trouble.>>[LAUGH]
>>But he offered me a research assistanship so
that decided what I was going to do. I was gonna work on
Surface Acoustic Wave Devices. So these are devices like if you put,
you have a piece of electric material and you put a transducers on which is
like an interdigitated pattern, you can give it an electrical signal and
launch an acoustic wave and then you could convert that acoustic wave back to an
electrical signal with another transducer. And one thing you can do is with
the overlap of fingers and things, you can build in filtering mechanisms, and there were ideas for building
convolvers and correlators with them. And Bob was looking at
stuttering zinc oxide and silicon, so the zinc oxide
would be the pairs electric so you could merge the acoustic wave device
with the silicon electronic control. So that’s what I did for my PhD. I didn’t see too much
future in those devices. So I didn’t look for a faculty position at that time because I knew I would have
to start probably in a new direction. And I didn’t know if I could do that or
didn’t know how long that would take. And there’s that tenure
clock that’s running. So I looked for a job. But the job market was
really bad at that time. So I had two offers. One was from Intel, so I could go back and
do the same thing I was doing. And I thought well I’ve just spent
three years in graduate school and now I’m just gonna go back and
do the same thing. And the other offer was Texas Instruments. But I probably got that because it had to
do with Surface Acoustic Wave Devices, an area I knew. It didn’t have that much of a future. So I took the TI job because
TI was different back then. And it was a big defense contractor,
they had a central research lab. So I was gonna work in
the central research lab, and it was on a contract from DARPA for these
surface acoustic wave memory devices. So I knew that I was not gonna
be able to spend a whole career on surface acoustic wave devices. So I was sitting at my desk one day and
I got a call Bob Gunshire, my advisor, and he said Purdue has bought these
two molecular-beam epitaxy systems. So I’m gonna be using one
to grow two-six materials. And there’s a lot of faculty
interested in gallium arsenide and they want someone to collaborate
who works on the material side and they knew the stuff you’d been doing with,
the sputtering the zinc oxide. So I came up for an interview,
got the job and been here ever since. So I joined in August of 1984,
got through the rank to four professors,
served two years as assistant dean, and I went through three dean
while I was assistant dean, so when I first started Henry Yang was
the dean but after a while he left to be chancellor at UC Santa Barbara
where he is still the chancellor. He’s been pretty successful there. Then, John Mclaughlin was interim dean for a while and
then Dick Schwartz became dean and maybe because of all
those transitions I just had enough with administrative work. So I went back to being, the best job at the university, being a full professor. Although later on I got talked into
being director of graduate admissions. Was the associate head at that time and Talked me into being
Graduate Admissions Director. And then,
when he became an Associate Dean, he talked me into being Associate Head. And when I took on that position
of Associate Head, I thought well, I’ll do it for at least two years,
maybe three years. And, as Roger Waters said, and then one day you find ten
years have got behind you.>>[LAUGH]
>>[INAUDIBLE]>>[LAUGH]>>So let me talk a little bit about my, Time and rank as a professor here. These are just the collaborators I’ve had
who have had faculty positions at Purdue. None of the students, a lot of students
and a lot of outside collaborators. In parentheses is the number
of co-authored publications, so I’ve had an enormous amount of
help from a lot of people here. The nice thing about Purdue is
it’s a very collaborative place throughout the university. So I’ve worked with people
in the school of science, electrical engineering,
chemical engineering, all over. So the first MBE, I spent probably 75% of my time always fixing it. I was able to squeeze out enough good results that we were able to raise
funding and then we got a superb machine. This very molecular beam epitaxy machine. So this thing allows you to grow crystals
of like, we were doing aluminum material. And have basically atomic layer control. Here’s, I actually don’t have
any many good pictures of it. But that’s it, and right across
the atrium here, used to be my lab. Okay so when I first got here,
Mark Lundstrom and I were graduate students at the same time. He graduated about a year before I did and
stayed on the faculty. By the time I came back from TI, he was
extremely well established, especially in the simulation of solar self [INAUDIBLE]
So he had this idea about collaborating
in an experimental project. Because, for the simulations, you need
to know material parameters very well. So it was a program to measure things
like minority care diffusivities. Lifetimes, band gaps. And so, he allowed me to be
a Co-PI in a contract we wrote and got funded by
the Solar Energy Research Institute. So it was $600,000 for three years. And back then, that was a huge contract. After three years we got it renewed, but
I learned, because we only spent $540,000, we didn’t get another $600,000, they just
added $504,000 to what we had remaining, so we made sure we always
spent it after that. Got renewed for one year in ’88,
then two years, and then three years. And we were up for renewal in 94 and
the Northridge earthquake hit. And that’s when I learned when there is
a site, it is called a National Disaster, where the money comes from. They tax existing agencies and
it was the turn for the Department of Energy to be taxed. So, there was no money for renewal. That’s kind of what ended these studies. Let me just give you
a flavor of some of them. One of the things we wanted
to get is minority carrier, diffusion coefficients and
intrinsic carrier concentration okay? And intrinsic care concentrations
will vary with doping, so both these will vary with doping. So we were building junction transistors, and you don’t really
need gain from the transistor to measure the product of the minority
care diffusion coefficient, and the square the effective,
intrinsic carrier concentration. So if you look at a Gummel plot,
here’s the base current and the collector current. So there’s no gain. The gain’s like 10 to the minus 3 or
something. But this is an ideal, n equal 1
current for the collector current. So the only unknown is this product of
the diffusion coefficient and N I squared. And for a lot of stimulations,
that’s all you need. But there are times when you want to know,
say, the diffusion coefficient by itself. So we get. Well, there’s [INAUDIBLE]. So we plotted out N I squared
D versus [INAUDIBLE] material. To pull out the minority carrier diffusion
link we built a time of flight experiment. So we would hit a P-N junction
with a femtosecond pulse and create a profile minority
carriage at the surface. And those profiles would
diffuse to the junction and then get separated and the voltage across
the device would do something like this. So here’s a plot in the middle is
a diffusion coefficient of 30 and the 2 lines are if you increase it or
decrease it by 15%. So a pretty sensitive measure to getting
that minority carry diffusion coefficient. Friction. Okay, so then we get to plot that out, measure that for P and
N type material as a function of doping. And then you can pull out
the intrinsic carrier concentrations. And they vary because as you add doping
your changing the crystal release, especially with high dopings
in the band gapped arrows. But you also, as doping goes up, and
material starts to be degenerate, so the firming level starts pushing
towards and into the bands, and you actually get effective
band gap widening. Okay and also, for people to believe
these numbers, we had to show we were going material comparable to what
everyone was a building solar cells with which was a process called
organic chemical definition. And the world record was 24.8%. And the cell we made by
molecular was 23.8%. So it was near state of the art. Material. Okay, so
let me talk about another collaboration. When I interviewed,
the first time I met Jim Cooper, and we really hit it off in the interview. And back then,
people were thinking that we were going to make high speed microprocessors
got a gallium arsenide. So Jim was working on dynamic memories. And he was looking at storing charge at
the aluminum, gallium arsenide gallium, arsenide interface,
which had like a three tenths ev barrier. But, I was working on these memory
correlators that were using reverse bias pn junction. For storage and [INAUDIBLE] so
we kind of put things together and started working on these
[INAUDIBLE] random access memories. And I’ll tell you in a minute how that
led to working on silicon carbide. But just briefly, it’s very similar
to silicon except you’re not using the semiconductor next to the offside. You have a PNP with the middle
end region floating. So if you put a negative
bias on the left-hand side, your forward bias in this
junction reverse bias in this so you’re gonna pull some carriers out
of the floating region in between. So when this goes back to ground,
you have a deficit of carriers. And the way you lose those is
through thermal generation which is a very slow process
especially in It’s not with the 1.42 even bandwidth. Okay, so we’re working on
this project on memories and. We were having funding from this strategic
defence initiative organization. [LAUGH] [INAUDIBLE]
>>So we were at a program review, and
it was being run by our program monitor, which was Max [INAUDIBLE] from
the office of naval research. Max is a Purdue grad back from in the 50s. But there was a group from [INAUDIBLE]
research there also presenting their work, and they were working on silicon carbide,
they were making these single crystal wafers that were maybe an inch
in diameter silicon carbide, but it’s band gap was over
three electron volts. And Max goes Right,
we take Purdue’s memory and make it in silicon carbide and
we’ll have a non-quality memory. So we struck up a collaboration with John
Palmer and Kevin Cartwright research and started to shift from
working on silicon carbide. And so
the idea you have your storage cell and then you would have
electronics to access it. And the interesting thing
about silicon carbide, the thought was you’d do things
very similar to silicon. Because if you oxidize it the carbon
leaves this CO or CO2, and you’re left behind with a thermal
oxidized silicon carbon wafer. So, we had to do a lot of work and
process developments. So you get a lot of what they did
in the early days with silicon, tried all sorts of
different parameters and measuring surface state densities and
finding the ideal things to where you’d go to build effective
MOS technology in silicon carbide. We built that memory with 64-bit one. We even built some CCDs. You might say why would we build a CCD? Well, remember this was
the Strategic Defense Initiative Office. And they wanted to find ICBMs
as they were being fired. And one of the things that happens
is the UV signature from an ICBM. They wanted a solar blind photo detector. Silicon carbides, there were over
three EV band gap, could look right into the sun and not really see it, but
see the missile coming out of the sun. But we’ve finally moved on to things that
were more practice than silicon carbide and that’s powered electronics, so we
built some high-powered Schottky Diodes. Here’s one with a good block
voltages over 5,000 volts. We made some of the first silicon
carbide DMOS power transistors. This first one could block 750 volts. UMOS, this one was about 1500 volts. And although I haven’t
participated in this work for awhile, it’s still going
on here at Purdue. So let me just talk briefly about
something I worked on that was a lot of fun,
I don’t know if it ever did or ever will have any practical application
but it was a fun material of science. So Molecular Beam Epitaxy is
basically a fancy thermal evaporator. So, you have these cells that contain
materials like gallium, aluminium, arsenic, and you heat them up and
start evaporating material. And, here you have the shutter, so
you can turn the beam of atoms on and off abruptly. Here’s the wafer you’re going
to be growing on it, typically, the growth temperature of
the wafer is about 600 degrees C. The gallium is sitting at about 900,
and the arsenic’s sublime to 280. So, every gallium atom that hits
the surface is gonna stick. And that arsenic atom is gonna stick only
if there’s a gallium atom to bond to. So, you’re almost guaranteed you’re gonna
get stoichiometric material with growth conditions like this. So at Lincoln Labs there was
a break in at a thermal couple or something, and the temperature
of the substrate went way down. And they said let’s just ramp
it back up and keep growing, and they got a good crystal doing that. So that seemed interesting so
I started to look at that. And so we grew some material where the growth temperature
was down to like 200 degrees C. Do an x-ray, it’s a good crystal. This is from the underlying substrate. This is from the epilayer we grew. And what this shows is you have
a good crystal and this spread is due to a strain in the crystal because of
the excess arsenic that’s all over inside. And you can tell from this split that we
were getting about 1% excess arsenic. And then we took this material and annealed it and
most of the strain disappeared. Okay, so the arsenic was going somewhere. So, Nobua Oatska who was a professor
here in materials engineering, who is a great transmission
electron microscopist looked at it. It found that all the arsenic
was precipitating out. And that the longer you anneal the sample,
the bigger and fewer the participants. It was an Oswald ripening
process that was going on. So what we found is that we
could grow any concentration and size of arsenic precipitates
by a combination of growth temperatures and
how much we anneal the samples. So this is one we grew where
we had gallium arsenide, aluminum gallium arsenide, showing
you get the precipitates everywhere,. But when we look at this it seems like, well there’s a lot more
at the interface here. Right on the gallium arsenide side. So we grew a structure where
we had 1% arsenic throughout, and then annealed it and you saw
precipitates forming everywhere but they’re starting to
preferentially form in these 100 angstrom gallium arsenide regions, and the
more you anneal, the more they would move. Okay, it was strictly that the gallium
arsenic bond is weaker than the aluminum arsenic bond. So it’s a lower energy state
if all the precipitates end up in the gallium arsenide. Another interesting thing
was the effect of doping. So here’s a structure where we had some
beryllium doping just in these regions, and silicone doping here. And you’d see that the arsenic
precipitates preferentially go to where there’s anti-doping, and
away from where they’re p-type doping. So an arsenic precipitates forms
a Schottky barrier in gallium arsenide. So if your in an N type region,
it’s gonna be negatively charged. So if you think of
an arsenic precipitated, a neo temperature getting emitted
from the arsenic precipitate, it’s gonna be positively charged and
coulombically attracted back. So that’s why there’s that preferential, precipitating on the silicon regio. Okay, one thing I didn’t have
in that list of things was that in July of 1999 I had a startup
company called OptoLynx. And we were making high-speed
photodetectors and photodetector arrays for
datacom and telecom. So when is the worst time to get into
the datacom and telecom business? It’s right before February of 2000,
when the Internet bubble burst. Okay, so we didn’t make it. But I ran across this quote
that’s really accurate, and I think it’s accurate for any start-up. You are never closer to
your greatest failure when you are at the moment
of your greatest success. Okay, and there were at least three
times with OptoLynx where that happened. I knew someone well from conferences
who was on the investment board at TRW. And TRW back then was
investing in companies and giving them Technology because
they found it more profitable to have stock in a startup than
make the product themselves. Then they could sell that stock for a lot more than they would
get from the products. And we had a term sheet from TRW. And I happened to be on vacation
in Wyoming when that happened, and back then cell service wasn’t so
good at places. And one of our investors here, sent a Can’t think of the word, he changed, tried to negotiate the terms
and sent them a revised, term sheet. And the way TRW viewed it,
we could say yes or no. And that was saying no. So that was the end. And when I talked to my friend, he said, the board was sitting there going
who do these guys think they are? And that’s how I kinda felt. Well, so that didn’t work. So then we got a Working with Cypress Semi Conductors, because Cypress
was getting into the data com module. And one of the things they told us was
that if they were going to use our parts, they would only do that
if they would buy us. And after the dot com thing was imploding, and we lost the deal with TRW,
it was like, yes, you can buy us. Then we got a call from them saying, well, Cypress is getting out
of the datacom business, were never really getting into it,
and we’re all looking for jobs.>>[LAUGH]
>>So that, ended that, then there was a third chance. MO-X bought ordered $80,000 worth of our photodetectors to qualify
in their datacom modules. That’s when I learned that
a purchase order isn’t a contract especially if you’re a small
company that can’t afford lawyers, maybe. They took delivery on only $20,000
worth of it and never any more and at one point when they place that order, they said well we’re gonna be needing a
quarter million dollars a month of these. Which would have been enough
to really sustain us. Okay, so that was the end of OptoLynx. So I had to you know,
I had grad–former students of mine, other students were in the company, so
I kind of felt real responsible and I kind of let my whole research
program here kind of dwindle. In fact, when I came back, I had one
graduate student left who was just about to finish, so at that time, Mark Lundstrom
and [INAUDIBLE] had two large centers starting, that had big education and
engagement programs in them. And the faculty member who was going to
run that just decided to leave Perdue. So they asked me,
would I be the director of that. And I thought, well yeah, that’d give me time to kind of figure
out what direction do I want to go. So I started to do that, and
let me, for time’s sake, we had a lot of programs,
like summer research interns, but let me just talk about
one of them that I did with Ron Rifenberger,
faculty member in physics. We were working on a nanotechnology
museum exhibit, and we did it under the EPICS program,
and so the things we did were it was going to have a Lego model of how
an atomic force microscope worked. There would be kiosks with animations and
hands on activities. And then it was eventually displayed
at the Children’s Museum in Oak Ridge, Tennessee. And now a lot of parts are over in
the atrium of the Burck building. So that was just the flier for
it that was distributed in Oak Ridge. There’s a picture of it in Oak Ridge. So this is our cam lever arm and
here’s the Lego landscape. I’ll show you how it works here. Just like in atomic force microscopes, you have a Lego landscape,
the thing comes down and there’s a mirror here that then gets
deflected when the surface hits. So there’s a laser beam reflecting
off this, and when it gets bent, it no longer hits a photodetector
on that column so you know when you’ve touched, so
you know the height of where you touched. So you can go through and scan the whole
surface just like an AFM does, and there you can see the laser
beam being deflected. And there are some of
the kiosks with the animations. The children playing with some
of the hands on activities. And we had a camera up in
the corner of the room so we could watch, took a photograph
like every 30 seconds, and then when school groups came and
visited we could get nice pictures of it. And Ron had a lot of nice AFM pictures
that we made and had them framed and you can see a lot of
those over in Burke now. And once we moved it to Burke, for a while we used to have a lot of
summer groups come through for tours. Okay so, I mentioned how I
became associate head, and one of the tasks of being associate head
is making all the teaching assignments. And as associate head I had
a complete release from teaching. But,what I found early on is it’s
really tough to make all these teaching assignments and
when I would get down to the very end and to get it over with I’d
assign myself to a class. But I like teaching so that’s fine. So one of the courses back 10 years ago
I occasionally had trouble staffing was 311 Electromagnetic fields. And I had taught it a couple times before,
but not often. And so
I started to teach it more regularly, and the students were always saying at the
beginning, God, I dread taking this class. They put it off to the last minute,
junior level course but they take it their senior year because they need it for
senior design the next semester. I was thinking this is the greatest
material, it’s fundamental to all aspects of electrical engineering, you
can understand how many things operate. So motors, rail guns, spectrometers,
breaking, transformers, batteries. So. I started to make little demonstrations. Because I thought if they can
picture these things, then they can. Develop mental models, because I think
they were all approaching this and learning it as some abstract math. They weren’t developing mental
models of these things. Cuz we have the right mental model. You don’t have to remember equations. Like Gauss’ law, you just think of charge,
flux coming off, close surface. All that flux goes through it,
you can count it. That’s equal to the charging side. You got Gauss’ Law. You can do that with all these equations. This is one course where you
can understand everything from the very beginning. So, I started to make these
little demonstrations, and then I actually started to video them. So I could put them up on blackboard, too. And I then started to
make additional videos. I would make short ones
explaining concepts. And I probably have about 50 of these
two to ten minute videos that I use when I teach, that I put up on Blackboard. And let me give you some examples. So this is an old electrostatic motor, one of the first motors in
the 1700’s that we looked at. So basically you have
some aluminum strips. You put a high voltage on this end,
and ground this end. So what happens is you pull electrons
off air molecules and they get repelled, deposited on the aluminum,
positive to positive, it rotates. And then they get, you have a positive then on this side
that pulls electrons off this one. And these things really have
a lot of renewed interest now for micro-electric mechanical systems. A lot of motors and micro-machines in
silicon are being built on these concepts. And, There. Okay, so I connect up the high voltage and
it should start spinning. Okay, so that’s an easy demonstration, that you can actually teach
a lot of things with. Okay, we’ll come to another
one here in a second. If you recall Faraday’s first motor, it’s
very similar in structure to this one. So this is a clear way of
seeing that idea across b. You have a magnet here,
current flowing through the wires. So the magnetic field lines like this,
idea across b. You get a real remoter out of it. So that’s very similar
to the one Faraday did, except he had his wires
sticking in mercury. Here’s a linear motor. So this is a bare copper wire. So the battery is two magnets,
and where the magnets touch, then you have the current flowing. So there’s just a current flowing where
the magnet is, and so this causes a magnetic field down the center of
the coil that pushes on the magnet. How a battery works or
how static electricity works. So you put two materials together. They have different
electrochemical potentials. So there’s gonna be a transfer charge
until the electrochemical potential goes to zero, okay a lot of people think, well it’s friction rubs these
electrons off, it’s not that at all. It’s the different in
electrochemical potential. The rubbing just enhances
a lot of contact, but after the charge moves it stops. So the way you sustain that
electrochemical potential difference and make a battery is, you put a solution
between it, like salt water. So, that’s the salt water
soaked piece of cardboard. And so
the salt water pulls the charge off so that you can sustain the current
flowing and so you get a voltage. Okay and then you can take these and
stack them just like Volta did and keep adding up the voltages. So then you connect some wires,
and a light emitting diode, and you have a battery. And so
I have that running all during class. Okay, so then talk about, you a have a curve going around in
the loop that’s a magnetic dipole moment. Fingers to your right hand in
the direction of the current flow from in the direction of
the magnetic dipole moment. You put that in a magnetic field, it’s gonna rotate till your thumb points
in the direction of the magnetic field. Now in an actual motor, you’d switch the direction of
current to sustain the rotation. This one doesn’t have a commutator so what I’ve done is I haven’t taken
the insulation off one side, so when M points in the direction of B,
M disappears because the current’s gone, momentum carries it around,
so it’ll keep rotating. Okay. Okay, so then you get to Faraday’s law. Actually I won’t show the experiment,
but you can do all the experiments of Faraday’s law and
show the terms in the equation. It’s pretty simple and straightforward. But then I also show how you utilize
that to build an AC generator. So I have a water bottle that I
have one continuous wire around, attached to a light emitting diode. There’s a magnet attached to this stick,
so you turn the magnet and you’re changing the magnetic
field inside the coil. And the faster you turn it,
the greater the change in magnetic flux, the greater the voltage
that’s gonna be generated. So start turning it here. Get it turning fast enough, you get
enough voltage for this to finally light. And it turns on and
off with the frequency of the rotation. Okay, and one last one. So we talk about currents
generating magnetic fields. If you have a coil, we’ve already gone over in class how that
looks very much like a bar magnet, okay. So then I have a coil on the bottom here,
okay so I’m feeding the audio output from this CD player, which the students
don’t know what a CD player is.>>[LAUGH]
>>And so as that varies the magnetic strength
of this coil is gonna vary, and, Okay, so I have a magnet here that then
I will pull close to this coil, and so there’s gonna be an attraction
between this coil, [MUSIC]>>[LAUGH]
>>[LAUGH] So I had to use Led Zeppelin. So there was a lot of sound, so
they could hear it in the classroom. Hey, I wanted to tell you about
something else I did in class. I started this exam point recovery, okay? So the idea is that the students
attend at least 80% of the classes. They had this opportunity, after each exam,
to recover lost points on one problem. So on that problem, they’d come and they
had 15 minutes not to rework the problem or similar problem, but
to teach me the concept behind it. So they really have to learn it a lot
better than they would’ve had to known to do it on the exam. Reason I did this was, first,
I like to have a full class. I think I teach better when
most of the students are there. And I hope that they’re coming,
they’re bound to learn something. And let me mention some of the benefits,
and a lot of these are ones
students have told me. They said, well, it clearly leads
to increased class attendance, cuz I can observe that one. And it’s an active learning by teaching,
cuz they have to come and teach me the concept. But some students have told me
it really reduces exam anxiety. Because if they get to
a problem they can’t do, they said, usually that just
ruins the whole exam for them. But now, they can say, I can spend
all my time on the other problems. I can get these points back. I can figure out this concept and
get the points back. And I usually have about ten problems,
so I don’t think it stops students from studying
because it’s not a lot to recover. But it’s enough that they’re gonna come
to class to have this opportunity. Okay, we think of exams as assessment,
but an exam is actually one of the best learning activities that we do. And there’s many aspects of that,
but I don’t have time to go into it. But that’s just one. Students become comfortable
presenting a technical topic. So some of them are pretty
nervous when they first come in, sit at a table with me and start talking. But it’s better that they’re nervous then
than when they’re on their job interview. And by the end they’re pretty good and
sometimes there’s a student who I can tell from what they put on
the exam, they were totally clueless. But they come in,
they’re an expert on the topic. So it’s a human thing, losses hurt. We’re more sensitive to losses than
gains by about a factor of two. That’s prospect theory. And so, they’ll put a lot
more time in studying to get ten points back than they would to earn
that ten points in the first place. Some of them start coming
to my office hours then, I guess it was enough incentive to get
them in now they feel comfortable. And if a student is
really totally lost I can help them at the beginning of the semester
rather than just fail them at the end. And as I said,
this is forcing them to learn concepts and not mathematical abstractions. And also the students
feel there’s the faculty member who really cares about them. Okay, the first time I did it, I didn’t
think I’d do it again cuz of the cost. With 55 students,
3 exams, 15 minutes each. I spent about 40 hours that
first semester doing this. But I thought the results were so
good that I kept doing it. But it could probably be something you
could incorporate and have your TA do. Just to give you some idea,
these are the times recently I taught 311 before I incorporated
this exam point recovery. The GPAs used to be about 2.7,
now they’re around 3.1, 3.2. Even though there’s really the number
of points being recovered, on average isn’t really that much. Sad thing is it really hasn’t helped this
category because there’s always a few students who just, no matter what,
they won’t come to class. And I think the students
really enjoy 311 now. They tell me that. And there’s something that
started happening a few years ago at the end of class.>>Okay, so I guess that’s it for
this semester. I really enjoyed the semester. I always like teaching this material and I
hope you enjoyed it and learned a lot too. And I’ll see you next week during
office hours and at the final.>>[APPLAUSE]>>Thank you.>>Okay so I’m already over time I think,
so I think I’ll stop here and see if anyone has any questions.>>[APPLAUSE]
>>Thank you.>>All right, questions?>>Where did you get
your experiments from?>>I just thought of some of them. Some of them you can find concepts
close to them on the web, or but I really wanted not to point them within
things and build them myself because one thing I found that most of the time I’ve
put something together it wouldn’t work. So actually we learned a lot by doing it. And so it’s something I
can stress to the students that we really don’t learn things
until you actually try to do them. Well, let’s thank Mike again.>>[APPLAUSE]